What makes a planet habitable?

Astron Astrophys Rev (2009) 17:181–249 DOI 10.1007/s00159-009-0019-z REVIEW ARTICLE

What makes a planet habitable? H. Lammer · J. H. Bredehöft · A. Coustenis · M. L. Khodachenko · L. Kaltenegger · O. Grasset · D. Prieur · F. Raulin · P. Ehrenfreund · M. Yamauchi · J.-E. Wahlund · J.-M. Grießmeier · G. Stangl · C. S. Cockell · Yu. N. Kulikov · J. L. Grenfell · H. Rauer

Received: 25 August 2008 / Revised: 2 February 2009 / Published online: 10 April 2009 © Springer-Verlag 2009

Abstract This work reviews factors which are important for the evolution of habitable Earth-like planets such as the effects of the host star dependent radiation and particle fluxes on the evolution of atmospheres and initial water inventories. We discuss the geodynamical and geophysical environments which are necessary for planets where plate tectonics remain active over geological time scales and for planets

H. Lammer (B) · M. L. Khodachenko · G. Stangl Space Research Institute, Austrian Academy of Sciences, Graz, Austria e-mail: [email protected] M. L. Khodachenko e-mail: [email protected] G. Stangl e-mail: [email protected] J. H. Bredehöft Institut für Angewandte und Physikalische Chemie, Universität Bremen, Bremen, Germany e-mail: [email protected] A. Coustenis Laboratoire d’Etudes Spatiales et d’Instrumentation en Astrophysique (LESIA), Observatoire de Meudon, Meudon, France e-mail: [email protected] L. Kaltenegger Harvard-Smithonian Center for Astrophysics, Cambridge, MA, USA e-mail: [email protected] O. Grasset UMR, Laboratoire de Planétologie et Géodynamique, Nantes, France e-mail: [email protected]

123

182

H. Lammer et al.

which evolve to one-plate planets. The discoveries of methane–ethane surface lakes on Saturn’s large moon Titan, subsurface water oceans or reservoirs inside the moons of Solar System gas giants such as Europa, Ganymede, Titan and Enceladus and more than 335 exoplanets, indicate that the classical definition of the habitable zone concept neglects more exotic habitats and may fail to be adequate for stars which are different from our Sun. A classification of four habitat types is proposed. Class I habitats represent bodies on which stellar and geophysical conditions allow Earth-analog planets to evolve so that complex multi-cellular life forms may originate. Class II habitats includes bodies on which life may evolve but due to stellar and geophysical conditions that are different from the class I habitats, the planets rather evolve toward Venus- or Mars-type worlds where complex life-forms may not develop. Class III habitats are planetary bodies where subsurface water oceans exist which interact directly D. Prieur Laboratoire de Microbiologie des Environnements Extrêmes, Université de Bretagne Occidentale, Brest, France e-mail: [email protected] F. Raulin Laboratoire Interuniversitaire des Systèmes Atmopshériques, CNRS & Universités Paris 12 et Paris 7, Créteil, France P. Ehrenfreund Astrobiology Group Leiden Institute of Chemistry, University of Leiden, Leiden, The Netherlands e-mail: [email protected] M. Yamauchi Swedish Institute for Space Physics, Kiruna, Sweden e-mail: [email protected] J.-E. Wahlund Swedish Institute for Space Physics, Uppsala, Sweden e-mail: [email protected] J.-M. Grießmeier ASTRON, Dwingeloo, The Netherlands e-mail: [email protected] C. S. Cockell CEPSAR, Open University, Milton Keynes, UK e-mail: [email protected] Yu. N. Kulikov Polar Geophysical Institute, Russian Academy of Sciences, Murmansk, Russian Federation e-mail: [email protected] J. L. Grenfell Zentrum für Astronomie und Astrophysik, Technische Universität Berlin, Berlin, Germany e-mail: [email protected] J. L. Grenfell · H. Rauer Deutsches Zentrum für Luft- und Raumfahrt, Institut für Planetenforschung, Berlin, Germany e-mail: [email protected]

123

What makes a planet habitable?

183

with a silicate-rich core, while class IV habitats have liquid water layers between two ice layers, or liquids above ice. Furthermore, we discuss from the present viewpoint how life may have originated on early Earth, the possibilities that life may evolve on such Earth-like bodies and how future space missions may discover manifestations of extraterrestrial life. Keywords Habitability · Origin of life · Terrestrial planets · Subsurface oceans · Atmosphere evolution · Earth-like exoplanets · Space weather · Astrobiology 1 Introduction: requirements for habitability Discussing the habitability of a celestial body is very closely connected to the discussion of life outside Earth. Life of course cannot ignore its connection to habitability, but the reverse is not automatically true. The word “habitable” means “suitable to live in (or on)”, and life cannot originate without habitability. But as an abandoned house might be perfectly suitable to live in (habitable) this does not necessarily mean there is somebody living in it. The question to ask in regard to celestial bodies is not only “Is it habitable?”, but also “Could life have originated and evolved there?” Besides the potential habitats in the Solar System more than 335 giant exoplanets have been detected1 to date. The current status of exoplanet characterization shows a surprisingly diverse set of mainly giant planets. Some of their properties have been measured using photons from the host star, a background star, or a mixture of the star and planet. These indirect techniques include radial velocity, micro-lensing, transits, and astrometry (Beaugé et al. 2008; Fridlund and Kaltenegger 2008; Rauer and Erikson 2008). Earth is until now the only example of a known habitable planet. Compared to other terrestrial planets in our Solar System, Earth is unique: it has liquid water on its surface, an atmosphere with a greenhouse effect that keeps its surface above freezing, and the right mass to maintain tectonics (e.g., Kasting et al. 1993). Earth orbits its host star—our Sun—within a region that is called the habitable zone (HZ)—the region where an Earth analog planet can maintain liquid water on its surface is shown in Fig. 1. The classical concept of the HZ was first proposed by Huang (1959, 1960) and has been modeled by several authors (e.g., Rasool and deBergh 1970; Hart 1979; Kasting et al. 1993). The differences in the calculations are the climatic constraints imposed on the limits of the HZ. In all cases the stellar habitable zone is a spherical shell around a main sequence star where a planet with an atmosphere can support liquid water at a given time. The width and distance of this shell depends on the stellar luminosity that evolves during the star’s lifetime. The continuously habitable zone (CHZ) has been introduced as the zone that remains habitable around a star during a given period of time (Hart 1978). Liquid water seems to be an important requirement for habitability. Liquid water has been pointed out as the best solvent for life to emerge and evolve in. Some of the important characteristics of liquid water as a solvent include: a large dipole moment, 1 http://www.exoplanet.eu.

123

184

H. Lammer et al.

Fig. 1 The HZ (upper panel) and the chemistry composition (lower panel) of an Earth-analog atmosphere as a function of distance from its host star. The dashed-dotted line represents the surface temperature of the planet and the dashed lines correspond to the inner edge of the HZ where the greenhouse conditions vaporize the whole water reservoir (adapted from Kaltenegger and Selsis 2007)

the capability to form hydrogen bonds, to stabilize macromolecules, to orient hydrophobic–hydrophilic molecules, etc. Water is an abundant compound in our galaxy, it can be found in different environments from cold dense molecular clouds to hot stellar atmospheres (e.g., Cernicharo and Crovisier 2005). Water is liquid at a large range of temperatures and pressures and it is a strong polar–nonpolar solvent. This dichotomy

123


185

Table 1 Limits for life as we know it (Cockell 1999) Parameters

Limits for life as we know it

Temperature (◦ C)

100

Acidity (pH)

>12

UV radiation

On early Earth 200–400 nm range of the 25% less luminous Sun reached the surface

Atmospheric composition

Pure CO2 can be tolerated by some organisms. High N2 will not prevent life

Water availability, liquid H2 O on the surface

Liquid H2 O should be present but some halophilic organisms live in high (4–5 M) NaCl

is essential for maintaining stable biomolecular and cellular structures (DesMarais et al. 2002) and there are a large number of organisms that are capable of living in water. The inner Edge of the HZ is the distance where runaway greenhouse conditions vaporize the whole water reservoir and, as a second effect, induce the photodissociation of water vapor and the loss of hydrogen to space, see (McKay and Stoker 1989; Kulikov et al. 2007; Lammer et al. 2008; Tian et al. 2008) for a detailed discussion. The outer edge of the HZ is the distance from the star where a maximum greenhouse effect fails to keep the surface of the planet above the freezing point, or the distance from the star where CO2 starts condensing (Kasting et al. 1993). On an Earth-like planet where the carbonate–silicate cycle is at work, the level of CO2 in the atmosphere depends on the orbital distance: CO2 is a trace gas close to the inner edge of the HZ but a major compound in the outer part of the HZ. Earth-like planets close to the inner edge are expected to have a water-rich atmosphere or to have lost their water reservoir to space. The classical HZ is defined for surface conditions only, chemolithotrophic life, a metabolism that does not depend on the stellar light, can still exist outside the HZ, thriving in the interior of the planet where liquid water is available. Such metabolisms (the ones known on Earth) do not produce O2 and relay on very limited sources of energy (compared to stellar light). They mainly catalyze reactions that would occur at a slower rate in purely abiotic conditions and they are thus not expected to modify a whole planetary environment in a detectable way. From our experience on Earth we know that if life once originated, it seems that it can adapt to all kinds of extreme environmental conditions (e.g., Rothschild and Mancinelli 2001) as shown in Table 1. The basic habitability requirements for life as we know are constrained by the following main factors: – a certain time span where a celestial body can accumulate enough building blocks necessary for the origin of life, – liquid water which is in contact with these building blocks,

123

186

H. Lammer et al.

– external and internal environmental conditions that allow liquid water to exist, on a celestial body over a time span necessary that allows life to evolve.

After life has originated, depending on the evolution of the planetary environment, it may evolve to form complex multi-cellular life forms or it may remain as microbial life which may adapt to extreme environmental planetary conditions. Under such considerations planets with eccentric orbits near or within the HZ may also be potential habitable worlds (Williams and Pollard 2002) due to climate stability in the long-term because the crucial parameter is the average stellar flux received by the planet during an entire orbit. Other possible exotic habitable environments, like extrasolar giant planets’ satellites (e.g., Europa- or Titan-like environments) located outside the classical HZ, have not been discussed as targets in the search for life with a terrestrial planet finding mission, because its related biomarkers will not be detectable remotely. Life that does not influence the atmosphere of its planet on a global scale (like on Earth) will not be remotely detectable. Therefore, such potential habitats can only be investigated in the Solar System with appropriate space missions. The HZ concept is still evolving (e.g., Segura and Kaltenegger 2008) as we learn more about planetary formation, evolution and improve radiative transfer and 3-D atmospheric models that allow more accurate calculations of a planet’s temperature profile, as well as the relation and interaction between the host star radiation and plasma environment and the planet. The question of what makes a planet habitable is much more complex than having a planet located at the right distance from its host star so that water can be liquid on its surface. Furthermore, various geophysical and geodynamical aspects, the radiation and the host stars plasma environment can influence the evolution of Earthtype planets and life if it originated (Scalo et al. 2007; Khodachenko et al. 2007b; Lammer et al. 2007). For instance, a subsurface ocean within the satellite of a gas giant may be habitable for some life form, albeit not necessarily Earth-analog life. On the other hand there may be many habitable exoplanets, but life could never evolve from microorganisms to complex multi-cellular life. With the discovery of planets beyond the Solar System and the search for life in exotic environments such as Mars, Europa, Titan and Enceladus, the various habitats need a separate definition. In this work we try to focus on aspects which are relevant to habitability but were so far not addressed in detail in studies related to HZs. In Sect. 2 we discuss the sources of life and its building blocks, which are ingredients of the planetary nebulae and the planetary environments which form out of them. As shown in Fig. 1, planets within the classical HZ orbit at different distances around different stellar spectral-type stars. Because planets evolve together with their host stars, we discuss the difference and expected consequences of the radiation and particle environment of F, G, K and M stars in Sect. 3. In Sect. 4 we focus on factors which are relevant for the evolution of habitable terrestrial planets, but are not discussed in detail in the classical HZ concept. These factors contain basic geophysical conditions in relation to their host stars, the role of plate tectonics, magnetic dynamos, ionosphere boundaries, induced mag-

123


187

netic fields, as well as the protection of planetary atmospheres against stellar plasma flows. In Sect. 5 we define four habitat classes. In class I the stellar and geophysical conditions of Earth-like (analog) planets where surface life as we know it can evolve are represented. Class II habitats will focus on planets where in the beginning life may evolve because these planets start an evolution similar to class I types but due to different stellar and geophysical conditions the planetary environments and life evolve different than on Earth. Class III habitats are planetary bodies with subsurface water layers in the form of subsurface oceans which interact with silicates (e.g., Europa). Finally, class IV habitats have liquid water layers between two ice layers or liquids above ice (e.g., Ganymede, Callisto, Enceladus, Titan-lakes, and “Ocean planets”). After discussing and defining these habitability classifications, in Sect. 6 we give an overview of the present knowledge how life may have originated on early Earth, and the possible characteristics of habitable planetary bodies which fall in the four habitat categories. Finally, in Sect. 7 we discuss future space missions which may be helpful to investigate the habitats addressed in this work. 2 Cosmo-chemical aspects for the building blocks of life Astronomical observations in the last decades have identified gaseous molecules and solids including carbon chains, aromatic hydrocarbons, carbonaceous grains and carbon-bearing ices that are observed in different galactic and extragalactic regions (Henning and Salama 1998). It is interesting to note that the carbon chemistry in different space environments proceeds to form many carbonaceous molecules—small and probably very large—which are common on Earth (Ehrenfreund and Spaans 2007). Where are those compounds formed and how do they evolve? 2.1 Interstellar clouds The interstellar medium (ISM) hosts an active carbon chemistry that is strongly influenced by environmental conditions such as density, temperature and radiation conditions (Wooden et al. 2004; Snow and McCall 2006). Table 2 lists the parameters for interstellar clouds. Molecular production is most effective in diffuse and dense interstellar clouds. The ISM comprises a few % of the mass of the galaxy. Its main constituent is H and He gas. However, submicron dust particles, mainly composed of silicates and carbonaceous material, are also present in small concentrations (∼1% relative to interstellar gas). Dust can adsorb ice mantles in cold interstellar regions. In the so-called dark interstellar clouds, characterized by high densities (see Table 2) active chemical pathways lead to complex carbon molecules in the gas phase and solid state (Ehrenfreund and Charnley 2000). Dark clouds offer a protected environment for the formation of larger molecules. Those regions experience a limited radiation field of ∼103 photons cm−2 s−1 induced by cosmic rays (Prasad and Tarafdar 1983). Icy surfaces of small dust grains offer a surface for catalytic reactions to form carbon bearing ice molecules such as CO, CO2 ,

123

188

H. Lammer et al.

Table 2 Phases of the interstellar medium (adapted from Wooden et al. 2004) ISM component

Designation

Temperature (K)

Density (cm−3 )

Hot ionized medium

Coronal gas

106

0.003

Warm ionized medium

Diffuse ionized gas

104

>10

Warm neutral medium

Intercloud HI

104

0.1

Atomic cold neutral medium

Diffuse clouds

100

10–100

Molecular cold neutral medium

Dark clouds molecular clouds dense clouds Protostellar cores

106

Molecular hot cores

CH3 OH and others (Gibb et al. 2004). Observations at infrared, radio, millimeter, and sub-millimeter frequencies show that a large variety of organic molecules are present in the dense interstellar gas (http://www.astrochemistry.net lists more than 150 molecules). Among them are simple molecules, such as CO, the most abundant carbon-containing species, with a ratio of CO/H2 ∼ 10−4 , but nitriles, aldehydes, alcohols and hydrocarbon chains have also been identified. Furthermore, Kuan et al. (2003) recently claimed that they detected the amino acid glycine in “hot-cores” that represent highdensity regions close to forming stars. The diffuse ISM is characterized by lower density and higher temperatures (more than 100 K). Ices are not present in those regions and a strong radiation field of ∼108 photons cm−2 s−1 (Mathis et al. 1983) dominates the formation and evolution of carbon compounds. Small carbonaceous molecules in the gas phase are easily destroyed by radiation. The identification of many small carbon chains in dense clouds implies that their destruction is well balanced by active formation routes. Circumstellar envelopes are regarded as the largest factories of carbon chemistry in space where molecular synthesis occurs on short timescales (several hundred years, Kwok 2004). Acetylene, a molecule formed in abundance in such regions, seems to be a key species that drives the chemistry towards aromatic rings, stable carbon molecules which are especially resistant to radiation. Cosmic abundances in the ISM are derived by measuring elemental abundances in stellar photospheres, the atmospheric layer just above the stellar surface. These cosmic elemental abundances determine the amount of elements available for the formation of molecules and particles. In dense interstellar clouds gaseous CO can account for ∼20% of the cosmic carbon. Figure 2 shows examples of carbonaceous matter present or anticipated in interstellar clouds. Several allotropes of carbon, among them diamonds, graphite and fullerenes have been identified in space environments, in particular in meteorites (Ehrenfreund and Charnley 2000; Pendleton and Allamandola 2002). Diamonds were proposed to be the carriers of the 3.4 and 3.5 mm emission bands observed in planetary nebulae. However, graphite is only present in meteorites and has not been identified in the ISM up to now. The polyhedral C60 fullerene geometry proposed by Kroto et al. (1985) is related to the presence of fullerene compounds in interstellar space. Until now fullerenes have

123


189

Fig. 2 Examples of carbonaceous matter present or anticipated in interstellar clouds. Several allotropes of carbon are found in space environment, e.g., in carbonaceous meteorites (diamonds, graphite, fullerenes, amorphous carbon). However, the presence of fullerenes and graphite in the interstellar medium is uncertain and the abundance of nano-diamonds is rather low. Aromatic hydrocarbons are the most abundant form of carbon in the universe. They exist as polycyclic aromatic hydrocarbons (PAHs) in the gas phase and are likely to be present in high abundance as aromatic networks (amorphous carbon) in carbonaceous grains

only been identified in meteorites (Becker and Bunch 1997). Possible fingerprints of the C+ 60 ion were discovered in the near-infrared spectra of diffuse interstellar clouds (Foing and Ehrenfreund 1994; Ehrenfreund et al. 2006). Cosmic dust models indicate that the majority of carbon in diffuse clouds (up to 80%) is incorporated into carbonaceous grains and gaseous polycyclic aromatic hydrocarbon (PAHs). PAHs are ubiquitous in galactic and extragalactic regions and are the most abundant carbonaceous gas phase molecules in space (Allamandola et al. 1999; Smith et al. 2007). There is spectroscopic evidence that carbonaceous grains are predominantly made of amorphous carbon (Mennella et al. 1998; Henning et al. 2004). Other forms such as C-onions and nanotubes have been proposed as well. It is a crucial goal for astrochemistry to identify the nature of this abundant aromatic material.

2.2 From interstellar clouds to forming planetary systems Interstellar clouds provide the raw material, namely gas and dust, for the formation of stars and planetary systems (Boss 2004). Astronomical observations of Solar System

123

190

H. Lammer et al.

objects in combination with the analysis of extraterrestrial material (meteorites and interplanetary dust particles IDPs) provided us with important insights into the processes that occurred during planetary formation in our solar nebula. Our Solar System was formed about 4.6 Gyr ago through the gravitational collapse of an interstellar cloud. Infalling interstellar gas and dust were thoroughly mixed and processed in order to form Solar System bodies. The small interstellar dust particles grew bigger and accreted more material, eventually forming planets or moons. The remaining material, not incorporated into planets, formed the populations of comets and asteroids that experienced a diverse history. A substantial fraction of small Solar System bodies impacted the young planets in the early history of the Solar System. The large quantities of extraterrestrial material delivered to young planetary surfaces during the heavy bombardment phase may have played a key role in life’s origin (Chyba and Sagan 1992). Recent data from the Stardust mission confirmed large-scale mixing in the solar nebula (Brownlee et al. 2006). The carbonaceous inventory of our Solar System is a mixture of materials including: – highly processed material that was exposed to high temperature and radiation; – newly formed compounds; – relatively pristine material with strong interstellar heritage. Future observations of organic compounds in our Solar System and the analysis of extraterrestrial material will provide us with more knowledge on the evolution of carbonaceous compounds and processes that occurred during the formation of our Solar System (Roush and Cruikshank 2004; Ehrenfreund et al. 2002). In the inner Solar System where terrestrial planets formed no original carbon material (volatile or refractory) could have survived the high temperatures. Organic matter on terrestrial planets must have formed after the planetary surfaces had cooled down. The extraterrestrial delivery of organic material during impacts of small Solar System bodies may have delivered important carbon compounds (Ehrenfreund et al. 2002). In order to reconstruct scenarios for the origin of life a prerequisite is to quantify the intrinsic inventory of carbonaceous material on Earth as well as the extraterrestrial influx from comets and meteorites. Another crucial question will be to investigate how this organic material could have assembled to form complex bio-molecules. Our universe provided carbon and dust in large amounts as early as 500 million years after the big bang (Spaans 2004). This indicates that the formation of planetary systems with rocky terrestrial planets may have occurred long before the formation of our own Solar System. Astronomical observations using better instrumentation allow us to reveal more details on the composition of interstellar dust and gas, comets and asteroids. These data and results from more sensitive techniques applied to the analysis of meteorites can determine more accurately the molecular inventory of material that was transported to young planets by exogenous delivery. From the latest stage of cosmochemical discoveries we can summarize that the building blocks of life should be available on every planetary system. After the planets have formed the molecules necessary for life are delivered by meteorites and comets to these early planets. Depending on the environmental conditions and the time-span where these conditions (e.g., water, surface temperature, surface atmosphere pressure,

123


191

etc.) remain non-destructive to the delivered molecules, life may origin like it did on early Earth. 3 Stellar activity, radiation and plasma environment of main sequence stars One of the major questions of current and future exoplanet finding missions (CoRoT, Kepler, SIM, Darwin/TPF-C/-I, etc.) is which of the main sequence star-types (M, K, G, F) may be good or at least preferred candidates for hosting habitable terrestrial planets? Obviously, the search for Earth-like exoplanets should not be limited only to the Sun-like G-type stars (e.g., Scalo et al. 2007; Kaltenegger et al. 2009). It needs to be extended also to lower mass M- and K-type stars, as well as to slightly more massive F-type stars. These stars have masses lower than 2.0 MSun , where MSun is the solar mass, and lifetimes longer than 1 Gyr. 95% of the stars in the mass range between 0.1 and 2.0 MSun , are M-type dwarf stars (M < 0.6MSun ), which are also the most numerous stars in the galaxy. This makes the study of the main sequence low mass stars and their potential influence on the planetary environments very important to the characterization of exoplanets and their possible habitability (Lammer 2007; Segura et al. 2003). The stellar X-ray and EUV (XUV) radiation and the stellar wind constitute a permanent forcing of the upper atmosphere of the planets, thereby affecting the atmospheric evolution and chances for life to emerge there (Kulikov et al. 2007; Lammer et al. 2008; Tian et al. 2008). The effect of these forcing terms is to ionize, heat, chemically modify, and slowly erode the upper atmosphere throughout the lifetime of a planet as shown in Fig. 3. The closer to the star the planet is, the more efficient are these processes. Protection of the upper atmosphere of a planet against stellar XUV/EUV and stellar wind factors requires a strong intrinsic dipole magnetic field (Khodachenko et al. 2007a,b). In general, the boundaries of the HZ shown in Fig. 1 change throughout the star’s lifetime as the stellar luminosity and activity evolve. This evolution is different for different star types, and furthermore depends on their age and the location

Fig. 3 Illustration on the radiation and plasma environment interactions with the upper atmosphere

123

192

H. Lammer et al.

of the HZ around the host star. This makes the type and age of a host star an important factor, which affects the evolution of potentially habitable worlds.

3.1 Radiation environment of lower mass stars Activity in late-type stars (i.e., spectral types G, K, M) has been the subject of intense studies for many years. The relevant physical phenomena of stellar activity and their observational manifestations include modulations of the stellar photospheric light due to stellar spots, intermittent and energetic flares, coronal mass ejections (CMEs), stellar cosmic rays, enhanced coronal X-rays, and enhanced chromospheric UV emission (see Ayres 1997; Gershberg 2005; Scalo et al. 2007 and references therein). Numerous observations of stars in clusters have revealed that single late-type stars spin down monotonically with their age which seems to roughly follow a power-law expression: Prot ∝ t −1/2 (e.g., Skumanich 1972; Newkirk 1980; Soderblom 1982; Ayres 1997) because of angular momentum loss. Earlier studies already pointed out a strong correlation between the rotation rate of a star and its activity level (Wilson 1966; Kraft 1967). This correlation shows interdependence between stellar activity and age (e.g., Keppens et al. 1995; Ribas et al. 2005). For solar-type stars this has been studied within the Sun in Time project (Ribas et al. 2005 and references therein). Based on a large number of multi-wavelength (X-ray, EUV, FUV, UV) observations of a homogeneous sample of single nearby G0–5 main sequence stars with known rotation periods and well-determined physical properties, luminosities and ages (covering 100 Myr–8.5 Gyr), it has been concluded that the solar integrated XUV flux (0.1–120 nm) was higher by a factor of six 3.5 Gyr ago than compared to the present state. Also, during the first 100 Myr after the Sun arrived at the zero-age main-sequence (ZAMS), the integrated XUV flux was up to 100 times more intense than today (Ribas et al. 2005). After this very active stage, the solar XUV flux quickly decreases with time following a power law relationship (t[Gyr]−1.72 ). In addition to that, the results of Wood et al. (2002, 2004, 2005) and Holzwarth and Jardine (2007) indicate that young solar-type stars may have massive stellar winds connected to their mass loss rates, 100–1,000 times that of the present solar value. Audard et al. (2000) found that the energy of flares correlates with the stellar activity (as characterized by the X-ray emission). The occurrence of energetic flares increases with increasing log(L x /L bol ) and then stays constant for saturated stars. Here, L x is X-ray luminosity of a star, which is used as a proxy of its overall activity, and L bol is bolometric luminosity. The time span of activity saturation is given in Table 3, which presents the evolution of log(L x /L bol ) with time for the stars of various masses, including solar-type and M-type stars (Scalo et al. 2007). This activity–age portrait also shows that lower mass stars spend more time in the saturated plateau before beginning their activity decay. In particular, solar-type stars stay at saturated emission levels until they age to ∼100 Myr, and then their X-ray luminosities rapidly decrease as a function of age following a power law relationship ∝ (t[Gyr])−1.72 (Ribas et al. 2005). M-type stars, on the other hand, have saturated emission periods up to 0.5–1 Gyr (and possibly longer for late M stars). After that the luminosity decreases in a way similar to solar-type stars.

123

What makes a planet habitable? Table 3 Time span in Gyr where L x /L bol(Sun) as a function of stars with masses ≤ 1MSun where the L x /L bol(Sun) is about 1,700 and ≥ 100 times larger than at the present Sun (after Scalo et al. 2007)

193 MSun

t [Gyr] for 1,700 L x /L bol(Sun)

t [Gyr] for ≥ 100L x /L bol(Sun)

1.0

∼0.05

∼0.3

0.9

∼0.1

∼0.48

0.8

∼0.15

∼0.65

0.7

∼0.2

∼1.0

0.6

∼0.3

∼1.47

0.5

∼0.5

∼2.0

0.4

∼0.75

∼3.0

0.3

∼1.0

∼4.15

0.2

∼1.58

∼6.5

0.1

∼4.6

>10.0

As a rule of thumb, the time evolution of M-type star X-ray luminosities could be considered as a constant luminosity L x ∼ 7 × 1028 erg s−1 up to the end of the saturation phase followed then by the power law decrease. K type stars remain at active emission levels somewhat longer than G stars and afterwards the activity decreases by following a power law relationship. Early M-type stars appear to stay at high activity levels until ages of about 1 Gyr, and then decrease in an analogous way to G and K-type stars (Scalo et al. 2007; and references therein). In general, early K-type stars and early M stars may have XUV emissions of about 3–4 times and about 10–100 times higher, respectively, than solar-type G stars of the same age (Ribas et al. 2005). In addition to that, partial or total tidal-locking of the terrestrial planets inside HZs of the low mass stars has long lasting consequences regarding plate-tectonics, generation of internal magnetic moments, and as a result the protection of the planetary atmosphere against erosion by stellar wind and CMEs. Weakening of the atmosphereprotective magnetosphere also results in easier penetration of galactic (Grießmeier et al. 2004, 2005, 2008) and stellar cosmic rays into the planetary atmosphere. This can lead to the kinds of ozone-depletion chemistry (Grenfell et al. 2007, 2008), similar to that on Earth in the polar regions, where energetic particles can more easily enter the upper atmosphere. If M star activity is accompanied by energetic particle emission, i.e., stellar cosmic rays, then the energetic particle flux due to the parent star could be orders of magnitude larger than that on the Earth. Similar effects can be expected for a tidally locked planet with a weak magnetic moment and therefore a stronger galactic cosmic ray exposure (Grießmeier et al. 2005, 2008).

3.2 Particle and plasma environment and habitability aspects 3.2.1 Stellar winds and Coronal Mass ejections Along with the high and long lasting stellar XUV emissions and energetic particle fluxes, another crucial factor of stellar activity which may strongly affect potential

123

194

H. Lammer et al.

habitability of a planet is the stellar wind plasma flow. The stellar plasma flow interacts with planetary magnetospheres, and in the case of weak planetary intrinsic magnetic fields, the stellar plasma flow can erode the planetary atmosphere and surface. Important parameters for characterization of the “stellar plasma–planetary atmosphere/surface interaction” are the stellar wind density and velocity, which are highly variable with stellar age and depend on the stellar spectral type as well as on the orbital distance of the planet. We currently have a good knowledge of these characteristics in the Sun, whereas the amount of data of this kind related to other stars is much more limited. While the plasma properties of the solar wind can be measured directly with satellites, the situation is more difficult for other stars. The dense stellar winds from hot stars, cool giants/supergiants and young T-Tauri stars produce spectroscopic features from which the wind parameters can be derived. The tenuous winds from solar-like stars, on the other hand, are a few orders of magnitude less massive and thus cannot be detected spectroscopically. There have been a number of claims of detection of mass flows in late-type stars, which however are not exempt of some controversy (e.g., Lim and White 1996). Different methods have been used to attempt direct measurements of mass loss rates in M-type stars, most notably using mm wavelength observations (e.g., van den Oord and Doyle 1997). Recently there have been important developments towards indirect detections of stellar winds through their interactions with the surrounding ISM. These observations were performed either in X-rays (Wargelin and Drake 2002) or in the strong Lyman-α emission feature (Wood et al. 2002, 2004, 2005). In particular, recently the stellar mass loss rates and related stellar winds were estimated for several nearby main sequence G- and K-type stars using Hubble Space Telescope (HST) high-resolution spectroscopic observations (Wood et al. 2002) and the analysis of the hydrogen Lyman-α absorption features. The HST observations by Wood et al. (2002, 2004, 2005) revealed neutral hydrogen absorption features associated with interaction between the stars’ fully ionized coronal winds and the partially ionized local ISM. From these absorption features the mass flux of the stellar wind as a function of stellar activity was modeled. Based on the study of a small sample of observed solar-like stars it was concluded that the mass loss and therefore the stellar wind mass flux decreases with stellar activity and stellar age. From these observations one can conclude that younger solar-like stars have much denser and faster stellar winds than the present Sun. A correlation between the mass loss rates and the X-ray surface flux indicates an average solar wind density of 100–1,000 times higher than today during the first hundred Myr after a solar-like star reaches the ZAMS (Wood et al. 2002, 2004, 2005). However, one should not forget that these results follow from observations of only a few K- and G-stars and to get a detailed knowledge of the stellar wind mass flux and mass loss of young stars more observations are needed in the future (Wood et al. 2005). Furthermore, it is known from observations of our Sun that active stars possess strong eruptions of coronal mass (e.g., CMEs), occurring sporadically and propagating as large-scale plasma-magnetic structures, which disturb the interplanetary space. CMEs appear as one of the important factors of the stellar activity. Traveling outward from the star at high speeds (up to thousands kilometers per second in the case of the Sun), CMEs create major disturbances in the interplanetary medium and produce

123

What makes a planet habitable? Table 4 Power-law approximated minimal and maximal CME plasma densities as function of distance from a Sun-like star (Khodachenko et al. 2007a,b)

195

d(AU)

n CME(min) (cm−3 )

n CME(max) (cm−3 )

0.005

∼106

∼108

0.01

∼105

∼107

0.05

∼104

∼105

0.1

∼103

∼104

0.2

∼100

∼500

0.5

∼50

∼70

1.0

∼7

∼10

strong effects on planetary environments and magnetospheres. Because of the relatively short range of propagation of the majority of CMEs, they should most strongly impact the magnetospheres and atmospheres of close orbit (0.3 AU) n CME is determined from the in-situ density measurements of magnetic clouds. Based on these data Khodachenko et al. (2007a,b) provided combined power-law approximations of CME −2.3 density decrease with the distance to a star (Table 4): n min ejecta (d) = 4.88 (d/d0 ) −3.0 , where d = 1 AU and d is taken in AU. Besides and n max 0 ejecta (d) = 7.10 (d/d0 ) that, the average mass of solar CMEs was estimated to ≈1015 g, whereas their average duration at distances (6–10) RSun is close to ≈8 h. Table 4 shows the estimated minimum and maximum CME proton densities as function of orbital distance. Until now, indications for stellar CME activity have mainly come from absorption features in the UV range during the impulsive phase of strong flare events, or from the blue-shifted components in time series spectra. Such features were found in the spectra of several active M-type stars, like EV Lac (Ambruster et al. 1986), AD Leo (Houdebine et al. 1990), and AU Mic (Cully et al. 1994) and were interpreted as indications of rapid mass ejections. In general, Cully et al. (1994) concluded that CMEs on those stars might be much stronger than the solar events. For example, Large enhancements in the far blue wings of the Hγ and Hδ lines during a strong flare on AD Leo, observed by Houdebine et al. (1990) were explained as a powerful CME-event with an estimated line-of-sight speed 5,800 km/s; 7.7 × 1014 kg mass and 5.0 × 1027 J kinetic energy. Whereas, modeling of the observed characteristics of a flare on AU Mic, detected with the Extreme Ultraviolet Explorer have led Cully et al. (1994) to propose a plasmid with the initial ejection speed of 1,200 km/s and mass 1017 kg. Discrete absorption features typical for a CME were also found in UV spectra of the eclipsing binary system V471 Tauri, consisting of a K-type dwarf star and a white dwarf. These features were interpreted by Mullan et al. (1989) as the signatures of a material cloud, ejected from to the K-dwarf and moving relative to it with velocities of 700–800 km/s, which are comparable to the solar CME speeds. Bond et al. (2001),

123

196

H. Lammer et al.

analyzing the absorption features in HST spectra of V471 Tau, estimated the frequency of CME-events of the K-type star component of the system as 100–500 CMEs per day, which is approximately 100 times more than the CME occurrence frequency of the present Sun. Khodachenko et al. (2007a) provided an estimation of the critical stellar CME production rate (CME occurrence frequency), for which (and higher) a close orbit planet will experience a continuous action of the stellar CMEs, so that the discrete collisions of a planet with CMEs can be replaced by a continuous action of the CME plasma flow. This corresponds to a situation when each CME collides with the planet during the time interval of the previous CME passage over the planet. The estimates gave the critical CME production rate of about 36 CME per day, which is not very much higher than the CME occurrence frequency of the present Sun (≥6 CME per day). Under the conditions of continuous CME flow action on a planet, the parameters of the stellar wind (density, speed, etc.) should be replaced by the parameters of the CME plasma, which will mean harder and more extreme conditions for the planetary environments than that in the case of a regular stellar wind. To summarize this section we would like to emphasize that the HZs of active flaring stars are exposed to intense stellar XUV irradiation, extreme particle and stellar wind conditions over rather long time periods. This radiation and particle exposure would strongly impact the atmospheres of terrestrial type planets in these HZs and may therefore significantly limit their range, as compared to those that result from the traditional HZ definition, which is based on the pure climatological approach.

4 Habitability relevant factors which are not considered in the classical habitable zone definition 4.1 Basic geophysical conditions Besides liquid water on the surface of a planet, a second characteristic of habitable planets is an atmosphere which is dense enough that it can stabilize the surface temperature of the planet through climate feedback like the “greenhouse effect” (e.g., Kasting et al. 1993). The greenhouse effect is caused by compounds that are very efficient absorbers in the infrared but not in the visible. The visible light of the parent star reaches the planetary surface and is remitted in the infrared where part of the energy is absorbed by the atmospheric greenhouse gases, increasing the temperature of the planet. Carbon dioxide (CO2 ), methane (CH4 ) and water (H2 O) raise the surface temperature on Earth by an average of 15◦ C, above the freezing point freezing. A planet has to accrete enough volatiles during its formation to have an atmosphere and it has to be massive enough to maintain this atmosphere. Venus and Mars demonstrate the limits of planetary habitability for life as we know it on Earth. Venus is the best example of what happens to a planet when the surface temperature exceeds a certain limit (like at the inner edge of the HZ). Venus has a surface pressure of about 90 bars and a surface temperature of about 480◦ . Present Venus has an atmosphere but only a tiny amount of water. It is uncertain if it did have water after its formation. If Venus had a similar water reservoir as Earth it was

123


197

converted to vapor due to the high surface temperatures of the planet. It remains an open question whether Venus lost its water due to a “runaway greenhouse” (Ingersoll 1969; Rasool and deBergh 1970; Walker et al. 1970) or a “moist greenhouse” (Kasting 1988, 1992) and thermal and non-thermal escape processes triggered by the young active Sun (Kulikov et al. 2006, 2007; Lammer et al. 2008). The “runaway greenhouse” occurs when water vapor increases the greenhouse effect, which in turn increases the surface temperature, leading to more water vapor that in turn heats the atmosphere. The other scenario is the “moist greenhouse” where water is lost once the stratosphere becomes wet but most of the water of the planet remains liquid. For both cases the loss of water happens high up in the atmosphere where water is photolyzed and H2 escapes while oxygen reacts with the planetary crust or is lost to space via non-thermal escape processes (Kulikov et al. 2006, 2007). Today Mars is a dry, frozen desert that cannot sustain life on its surface. The Martian atmosphere was lost and became too thin to warm the planetary surface. A minimum planetary size for a terrestrial planet seems to be a crucial factor for an Earth-analog habitat. About >4–4.5 Gyrs ago Mars may have had an atmosphere thick enough to maintain liquid water on the surface (e.g., Kulikov et al. 2007 and references therein) which led to the hypothesis of life on early Mars. Large impactors evaporated the Martian atmosphere and the low gravity of the planet was not able to retain the gas in the hot plumes created by those impactors (Melosh and Vickery 1989; Manning et al. 2006; Pham et al. 2009).

4.2 Planets with and without plate tectonics The most plausible model is that small bodies very quickly develop a stable thick lithosphere within the first 100 Myr (e.g., Nelson 2004), resulting in a smaller and weaker heat flow and therefore smaller convection cells. Liquid magma is still present within smaller bodies, which is shown by the fact that the maria at the Moon caused by large impacts are covered by lava after the late heavy bombardment (LHB) and, e.g., Olympus Mons on Mars. No other planetary body in the Solar System than Earth seems to have developed a system of continents or cratons, light-weight parts of the crust which do not vanish by subduction. Sediments are produced, which result in the first gneisses and granites, probably already before the LHB. The oldest formations are dated to about 4 Gyr ago, but some zircons are dated earlier and must have been formed about 4.4 Gyr ago (Nelson 2004). If the separation of continental crust on Earth started so early, all other terrestrial bodies should be expected to have experienced such a phase too. If started, plate tectonics will produce continental crust at subduction zones which are added to the cratons over time. Another model postulates that continents have been in a balance of accretion and constructions since 4 Gyr (Arndt 2004). Probably there are three stages: – a magma ocean 4.5–4.0 Gyr ago, – a transition phase with starting plate tectonics but heavy crustal breaks by superplumes 4.0–3.0 Gyr ago,

123

198

H. Lammer et al.

– since then the development of slow plate tectonics up to the present velocities (Eriksson and Catuneanu 2004). It seems that this process producing terranes, increasing continents by lateral accretion of pieces of the crust, and formatting of supercontinents by driving cratons together, has a period of a few 100 Myr (on Earth Vaalbara, Kenorland, Columbia, Rodinia, Pangaea). If there were plate tectonics on all terrestrial bodies, it seems likely that it stopped very early on small bodies, at least about 3.5 Gyr ago, because the LHB impacts are still seen as mentioned before. On Earth the first signs of life are detectable at that time (e.g., Schopf 2004). The picture at Venus seems to be more complicated. The two bodies of Earth’s size in the Solar System, Venus and Earth, differ heavily in the number of impact craters. The approximately 100 craters at Earth are much less than the approximately 900 on Venus. Even if some on Earth vanished by erosion processes this means that tectonic processes extinguished most of the craters on Earth, but not on Venus. The question is why Venus did not develop active plate tectonics similar to the Earth? One factor might be the high surface temperature which weakens the heat flow and slows down the convection (Valencia et al. 2007). Despite about 900 craters (http://www.lpi.usra.edu/resources/vc/vchome.shtml) from impacts, the landscape of Venus is dominated by lava flows and probably other formation processes. The sharp decline of Ishtar Terra to the West and the much softer one to the East may point to a subduction zone in the West caused by plate tectonics (Ansan et al. 1996; Lenardic et al. 1991; Vorder Bruegge and Head 1990; Janle and Jannsen 1984). In other regions plumes and superplumes might form large igneous provinces (LIP), providing heat transfer and producing surface structures (Nijman and de Vries 2004) like plateaus and volcanoes (hot spots). Plate tectonics should have stopped at Venus after the water inventory was lost (Nelson 2004). Without any surface probes the geological history of Venus cannot be deciphered. Plate tectonics together with permanent liquid water most likely provided the first environment for the evolution of life, the black smokers. Plate tectonics does not only regulate the composition of a terrestrial atmosphere by the cycling of volatiles, including the greenhouse gas CO2 and hence the surface temperature and planet habitability (e.g., Sundquist 1993; Kasting et al. 1993; Franck et al. 2000; Wolstencroft and Raven 2002), but it is driving evolution by always changing the environment. The chances for being created and being forced to evolve (and probably also for being destroyed) are better on an active planet than on a stable one. As illustrated in Fig. 4 plate tectonics is also an important factor for the generation of an Earth-like long-time strong intrinsic planetary magnetic dynamo, which protects the atmosphere from solar wind erosion and deflects high energy cosmic rays (e.g., Ward and Brownlee 2000; Grießmeier et al. 2005; Khodachenko et al. 2007b). Although the driving mechanisms for plate tectonics are not fully understood, the minimum requirements are a sufficient mass relevant for the heat flow to drive mantle convection, and water to lubricate plate motion (e.g., Regenauer-Lieb et al. 2001; Solomatov 2004). Water is the lubricant that allows the plates of the crust to slide and subduct, without water in the mantle the evolution of the planetary mantle and planetary tectonic engine would stop. Water makes the lithosphere deformable enough for

123


199

Fig. 4 Illustration of the connection between plate tectonics, mantle cooling and magnetic Earth-type dynamos. Upper panel Active geodynamic conditions due to plate tectonics. Lower panel Inefficient cooling in the mantle yields a one-plate planet with non or weak magnetic protection (courtesy of D. Breuer)

subduction of the crust to occur and it reduces the activation energy for creeping and the solidus temperature of mantle rock, thereby enhancing the cooling of the interior and the efficiency of volcanic activity. Large water reservoirs in the mantle and on the surface are interacting. The mantle loses water and other volatiles like CO2 through volcanic activity, and therefore helps to sustain the atmosphere. On the other hand, water and CO2 are recycled together with the subducting crustal rock (e.g., Breuer et al. 1996, 1997). Furthermore, the recycling of the crust through plate tectonics keeps the crust thin, which seems to be mandatory for plate tectonics to operate. If the crust is too thick, the lithospheric plate comprising the crust will be too buoyant to be subducted because on Earth the cratons became never subducted anymore after formation. Finally, plate tectonics seem to help to establish the right temperature conditions in the interior that are required to maintain the action of a strong magnetic dynamo for several billion years.

123

200

H. Lammer et al.

For terrestrial exoplanets the conclusion can be drawn that planets with sizes less than Earth or Venus (e.g., Martian sized bodies) lose their ability for plate tectonics very quickly. Water-rich Earth-sized bodies should maintain the convection cycle for a long time, thus forming dry continents and basins with water. Concerning “super-Earths” which are terrestrial exoplanets with sizes between R⊕ and 2 R⊕ the kinematic energy of the convection cells should be higher than on Earth. Plate velocities should be faster and orogenesis much heavier, resulting in higher mountains and more volcanic activity. It was shown by Valencia et al. (2007) that the crust on “super-Earths” which is comparable to the “oceanic crust” on Earth is even slightly thinner and the convection cells are only slightly larger, but plate velocities will increase approximately linearly with planet mass. However, the conclusion drawn from this, namely that mountains will not be higher and trenches not deeper seems to be questionable (Aguilar 2008). A faster plate velocity will result in a higher kinematic energy which produces orogenesis at subduction and collision zones. Plate tectonics on “super-Earths” depends on several factors, like the original water contents, the forming of cratons and the distribution of convection cells, all probably individual for each planet. Even on Earth variations are frequent within space and time and it is difficult to find a precise model to describe the processes leading from the magma ocean to the present day continents.

4.3 Magnetic dynamos and their role in atmospheric protection against solar/stellar plasmas Strong intrinsic planetary magnetic fields are generated by a complex interaction of hydromagnetic processes. The source of the internal magnetic field is the motion of a highly conductive fluid within the planet (i.e., the liquid outer core for terrestrial planets, or a region of electrically conducting hydrogen for gas giants). These large-scale motions can be produced by two different mechanisms that are believed to appear subsequently during the evolution of a terrestrial planet (Stevenson 1983, 2003; Stevenson et al. 1983), namely thermal and compositional convection. Large amounts of water and related plate tectonics could have played a role in keeping a magnetic dynamo alive over geological time scales. Depending on the initial water inventory of early Venus, possible plate tectonic activity and the related thermal history of Venus’ interior, theories of dynamo generation suggest that there could have been a magnetic moment on Venus of the same order as that of present Earth for about the first billion years after the origin of the planet (Stevenson et al. 1983). During that time, thermal convection from the heat left over from accretion could have driven the dynamo. After that this energy source diminished and no other source replaced it. The solid core formation in Earth’s interior keeps its dynamo working to this day by virtue of the related stirring of the molten layer around it. Present Venus may lack the necessary internal chemical or physical ingredients for solid core formation, or such processes may have stopped at an earlier time (Stevenson et al. 1983). Because the solidification of the inner core is thought to be the energy source for the present day terrestrial magnetic field, and smaller bodies thermally evolve more

123


201

rapidly than larger bodies, one can conclude that the terrestrial planets today are in three different magnetic phases, where Venus is most likely in a predynamo phase, not having cooled to the point of core solidification (Russel 1993). The Earth is in a dynamo phase and Mars is in a post-dynamo phase. Maintaining an intrinsic magnetic moment would have been problematic following the loss of water. Present Mars lacks a detectable global magnetic field (Acuña et al. 1998, 2001; Connerney et al. 2001). On the other hand, at small spatial scales the Martian crust shows remnant magnetism, which is, on average, about 10 times more magnetized than that of the Earth (Connerney et al. 2004). This suggests that Mars once had an active dynamo and the crustal remnants were acquired either by thermoremnant magnetization or by another process like chemical remnant magnetization. Acuña et al. (2001) proposed that Mars’ magnetic dynamo ceased operation relatively early in the planet’s evolution. These authors argued that large areas above the huge impact basins—Hellas, Argyre, Isidis lack measurable crustal fields, as did much of the impact basins in the northern lowlands. If these impact basins had originated in the presence of a strong ambient magnetic field, they would most likely have been magnetized as the rocks cooled below the Curie temperature, because a reworking process or shock demagnetization of the entire crust in the vicinity of these impact basins is very unlikely. Since the formation of these large impact basins is believed to have occurred at the end of the LHB ∼3.8 Gyr ago, one can conclude that the Martian dynamo ceased its operation after a few hundred Myr during the Noachian epoch. This hypothesis is in agreement with the model of Breuer and Spohn (2003) which predicts thermal convection early in the Martian history, so that the dynamo action could occur for hundreds of Myr. On the other hand Schubert et al. (2000) and Stevenson (2001) proposed that these large impact basins originated before the onset of the dynamo, because the magnetization of the southern highlands could also result from localized heating and cooling events that occurred after large impacts and basin formation. Further, Schubert et al. (2000) note that an early onset and cessation of the Martian dynamo is difficult to reconcile with the hypothesis of a dynamo driven by solidification of an inner core as it is thought to have occurred on Earth. Phase transitions in the Martian mantle and their interaction with the mantle flow may provide an alternative explanation for a late onset of a purely thermal dynamo (Spohn et al. 1998). They discussed how a putative layer of the mineral perovskite above a small and Fe-rich core could affect the thermal history of the core evolution. According to 2-D and 3-D convection model calculations by Breuer et al. (1997), a 100–200 km thick perovskite layer reduces the heat flow from the core since it forms an additional convective layer between the overlaying mantle and the core. A perovskite layer (Spohn et al. 1998) could thus be responsible for a late onset of the Martian dynamo, or a re-birth of the dynamo at a later time (Lillis et al. 2006). Thus, the models of the core evolution are consistent with both an early magnetic field and or a late onset of it (Connerney et al. 2004; Lillis et al. 2006). However, these models cannot predict the strength of the magnetic field. Unfortunately, there is as yet no observational constraint on the timing of the Martian dynamo. An early Martian dynamo with a minimum and maximum expected magnetic moment between 0.1 and 10 times that of the present Earth can be expected (Schubert and Spohn 1990).

123

202

H. Lammer et al.

4.3.1 The role of rotation and magnetic dynamos As discussed above, the magnetic field protects a planetary atmosphere against strong erosion by the solar or stellar wind and CMEs. Planets that are not protected by sufficiently strong magnetic shields and exposed to extreme stellar radiation and plasma flows can be rendered inhabitable due to non-thermal atmospheric erosion processes. Because planetary rotation and magnetic moment are linked, rotation is probably an important factor for planetary habitability. This is especially relevant for terrestrial planets in the HZ around M dwarfs. These planets are located so close to the star that they are tidally locked, i.e., have strongly reduced rotation rates (e.g., Grießmeier et al. 2005, 2008; Khodachenko et al. 2007b; Lammer et al. 2007). In many cases, the corresponding weak magnetic moment may not be sufficient to protect the atmosphere, leading to planets which may lose their atmosphere and water inventories even when they are located within the liquid water HZ. The generation of planetary magnetic moments is based on convective motion in the planetary core. As a requirement for convection to occur, the Coriolis force has to have a large effect on the flow. This condition, however, is easily satisfied, even for slow rotators like Venus (Stevenson 1983, 2003). Thus, when we discuss the influence of rotation on the magnetic dynamo further below, we always assume that rotation is strong enough for a dynamo to exist in the first place.

4.3.2 Dynamo models with rotational influence Under the assumption that the planetary magnetic field can be derived from a dipole moment (i.e., no higher multipoles), one can attempt to estimate the planetary magnetic dipole moment M from characteristic values of the planet using simple analytical models. This usually leads to simple scaling relations, which can be used to calculate the order of magnitude of the planetary magnetic dipole moment (or its surface magnetic field, which is proportional to Mr −3 , where r is the planetary radius). Several such models are summarized in Grießmeier et al. (2005). All these models yield increasing magnetic moments with increasing planetary rotation rates. Thus, the shorter the planetary rotation period is (i.e., the faster its rotation), the larger the resulting magnetic moment can be expected to be.

4.3.3 Dynamo models without rotational influence While the analytical models considered above predict an increase of magnetic moment with increasing rotation rate, numerical experiments indicate that the magnetic moment may be independent of the angular frequency. Christensen and Aubert (2006) and Olson and Christensen (2006) have studied numerically an extensive set of dynamo models in rotating spherical shells, varying all relevant control parameters by at least two orders of magnitude. These simulations indicate that the magnetic field is basically controlled by the buoyancy flux, and rotation does not seem to have any influence. This result is in contradiction with the analytical models discussed above.

123


203

To estimate the maximum impact on an atmosphere we assume that rotation has an influence on the planetary magnetic moment. Further studies are needed to clarify the relation between planetary rotation and magnetic field strength. 4.3.4 Estimation of magnetic moments for terrestrial planets with different size, mass and rotation periods Based on the formulas given by Grießmeier et al. (2005) one can calculate the magnetic moments for different situations. To illustrating the possible effect of the rotation on the strength of a magnetic dynamo we investigate a “super-Earth” and an “Earth-like” planet. We assume the following planetary structure (Cain et al. 1995; Léger et al. 2004): – the radius for the massive terrestrial planet of 6 M⊕ is about 1.63 r⊕ , – the core density is 10.6 103 kg/m3 for the Earth-like, and 15.5 103 kg/m3 for the large/massive planet case, – the size of the planetary core is rc = 0.55 r⊕ for the Earth-like, and rc = 0.52 r⊕ for the large terrestrial planet case and – the conductivity σ is roughly the same as on Earth for both cases. The resulting magnetic moment as a function of the planetary rotation period is given in units of the Earth’s current magnetic moment (M⊕ = 8 1022 Am2 ). The shaded areas in Fig. 5 represent the ranges of results obtained by the different magnetic moment scaling laws. For a rotation period of 10 h or more, one finds that Earth-like terrestrial planets can have a maximum magnetic moment of 2.5 M⊕ . Tidally locked planets have rotation periods equal to their orbital period (i.e., rotation periods 10–100 times larger than for the Earth), and are likely to have a much lower magnetic moment than the Earth (more than one order of magnitude lower).

Fig. 5 Ranges of expected planetary magnetic dipole moments as a function of rotation period and planetary mass. Light grey: Earth-like exoplanets. Dark grey: Super-Earths with 6 Earth masses and 1.63 Earth radii (Léger et al. 2004). Dotted lines indicate the current value of the Earth (rotation period and magnetic moment)

123

204

H. Lammer et al.

Large terrestrial planets of 6 Earth masses are likely to have a magnetic moment approximately a factor of 5 higher than Earth-like planets. As the planetary radius is also larger (1.63 r⊕ ), this corresponds to a surface magnetic field approximately 20–30% higher than on an Earth-like planet. For this reason, the magnetic shielding of a large planet is slightly better than that of an Earth-sized planet (assuming identical rotation rates). The improved magnetic shielding with respect to atmospheric erosion is presented in Sect. 4.3.5. The temporal evolution of the magnetic moment of an Earth-like planet caused by the temporal evolution of the rotation rate is discussed by Dehant et al. (2007). 4.3.5 Magnetic protection of slow rotating planets Planets with a small magnetic moment also have a small magnetosphere. This is quantitatively discussed for terrestrial exoplanets in Grießmeier et al. (2005, 2008) and Khodachenko et al. (2007b). For a small magnetic moment and/or a strong enough plasma flow (either stellar wind particles or solar like stellar CMEs), the magnetosphere is so small that its boundary surface (the magnetopause) is located below the outermost layers of the planetary atmosphere (Khodachenko et al. 2007b; Lammer et al. 2007). The results shown in Fig. 6 indicate that one can expect many planets within the HZ of dwarf stars with Venus- or Mars-like stellar plasma-atmosphere interactions. However, we have to keep in mind that dynamos and the resulting strong intrinsic magnetic fields, as on Earth, are only generated if the planet is geophysically active. If plate tectonics stop or do not work well on the planet, the magnetic dynamo would also stop working. As discussed in Sects. 4.2 and 4.3, as a result one obtains Venusor Mars-like unprotected planetary environments where the ionized particles within the upper atmosphere form a planetary obstacle.

Fig. 6 Region where, according to the estimations of Sect. 4.3.4, the magnetopause lies close to the planetary surface. Grey-shaded area the classical liquid water HZ as discussed in Sect. 1 (e.g., Kasting et al. 1993). Lightly dotted area: region where the magnetopause can be compressed to altitudes